You are on page 1of 3

International Journal of Computer Science Trends and Technology (IJCST) Volume 2 Issue 1, Jan-Feb 2014

ISSN: 2347-8578 www.ijcstjournal.org Page 49



Survey on localization of the Sensor nodes in Wireless Sensor
Networks
B.Dineshbabu
1
, T.Thirunavukarasu
2
Information Technology, SNS college of Technology, Coimbatore
Anna University, Tamil Nadu-India

ABSTRACT
The Wireless Sensor Networks is a group of nodes scattered in the network. Each node has certain processing capacity,
memory. It shares the resources and information among the nodes. The example of these services includes mobile services,
traffic monitoring, natural calamities, smoke detection and vehicle tracking applications. Sensor networks provides the fast
sensing and make necessary actions to provide reliable transfer of information. The sensor node helps to provide the better
gathering of data and it provides a backup through which the failure of one node is compensate by the other node. Sensor nodes
require less energy, battery power, reduced cost easy node deployment.
Keywords- Localization Algorithm, Jamming nodes, Multi-alteration


I. INTRODUCTION
Sensor Node Localization is the main process in the
Wireless Sensor Networks (WSN). Node localization will
improve the performance of WSN. To find out the location of
the sensor node is the main task, wrong prediction of the
location due to attackers will leads to many problematic
situations. An attacker can spoof by using another node
identity to make an attack which results in the compromise of
node and hence finding the location of the attacker will be
crucial task. We can find out the location of the sensor node
by either two or three dimensional through which the
landmark of the node can be identified. One of the main
problems is the coverage area where it becomes difficult to
cover all the areas in the network. Hence a survey is needed to
be taken for localizing the sensor node.

II. MACHINE LEARNING APPROACH IN
LOCALIZATION
In this paper [1] there are various techniques implemented
to find out the location of the sensor nodes, machine learning
is the different technique used for the localization. The
information which is shared by giving the signal strength as
the input.SVM is the machine learning approach for
classification by getting the input and by having the training
sets available, it predicting the corresponding output by
accurately predict the location of the sensor node and
determine the class of the sensor node. These training methods
include only the beacon nodes through which any sensor can
find the location without the knowing about other sensors.
Exact location based on the size of the training data that we
have already gathered. By combining the past and the current
information gathered we can find out the location accurately.
This will reduce the cost and need for every sensor to run the
localization algorithm. The important processes to find out the

III. JAMMER LOCALIZATION IN
WIRELESS SENSOR NETWORKS
In this paper [2] is to find the position of the jamming
device is a critical process so as to take the necessary actions
in order to the jammer in the network communication. To
provide an efficient algorithm called Double Circle
Localization (DCL). To deploy DCL in different situations
and make contrast with the present jamming localization
algorithms by simulation, by comparing with the other
methodologies it provides the exactness in localization of
jammer. This attack affect the network by stopping the
transmitters from transferring the data which results in packet
collisions even though the alternatives are choose to counter
these attacks like channel surfing. Advanced devices and high
standard protocols are required to prevent this attack. To make
effective strategies location of the jammer is essential. All
nodes are randomly deployed and there is no change in their
locations. This device always emit the electromagnetic power
by that it can sense the signal and stopping or preventing the
data which is transferred. This jamming signal will disturb the
communications in the sensor networks. The energy of the
jamming will be high to the nodes which are nearer to that
jamming node. The nodes which are jammed are to transmit
these jammed messages to the nearest node. With the help of
the boundary nodes we can detect the presence of the jamming
node and then the localization algorithm for the jamming
nodes are activated. To measure the landmark of the node can
be done by the Received Signal Strength (RSS) and also
provide the exact distance between the jamming nodes and
nearest nodes. To transfer the node from the predicted location
to the jamming location Virtual Force Iterative
RESEARCH ARTICLE OPEN ACCESS
International Journal of Computer Science Trends and Technology (IJCST) Volume 2 Issue 1, Jan-Feb 2014
ISSN: 2347-8578 www.ijcstjournal.org Page 50
Localization(VFIL) to run many times as pull and push by the
mutual virtual forces. When the nodes are outside the
predicted region it will make the jammed region near to it.
And the inside the region will push the unaffected nodes.
Centroid Localization will gather the location information
about the neighbouring node, weighted centroid localization
combine the information of the target node to provide the
accurate Jamming node Localization.

IV. ATTACK DETECTION IN WIRELESS
LOCALISATION
In this paper [3], Non-Cryptographic attacks cannot be
controlled by the traditional security services and hence it is
essential to detect the attack in Wireless Sensor Networks
(WSN).At first with the help Statistical Significance Testing
attack detection problem can be formulated theoretically. With
the help of localization approaches such as Multi-alteration
and signal strength to provide solutions for high detection rate
for attack detection. When there is any physical attacks
localization algorithm performance decreases. The output of
the compromised node make a serious risk by affecting the
applications, Most used localization approach is the
Multialteration which uses the techniques called Least
Squares(LQ) to make localization process. Many algorithms
use the signal strength to detect the attackers. By measuring
the physical properties we can calculate the distance between
them with the help of the Range based algorithms. Attack can
be detected with the help of signal strength. Attacker can
modify the signal strength and change the location of the node
with the help of low cost electronic devices. Certain
performance metrics can be used to make the high attack
detection range. Cumulative Distribution Function (CDF),
Detection rate (DR) are the performance metrics used for the
attack detection. The RSS has the ability to reuse the wireless
network infrastructure, the main benefit in the use of this
technique is that it does not depend on the localization
algorithms and it performs earlier than the localization process.
By selecting the correct threshold to prevent the false
detections, Detection of attack in signal authentication
provides better performance than the signal attenuation due to
its higher detection range when the signal is amplified by the
attacker. To make the highest detection range the Area Based
Probability (ABP) and Bayesian Networks (BN) which makes
the signal strength to further increase the detection range and
the location of node can be identified .Bayesian Network
provides the probability distribution of finding the unknown
locations with the help of Monte-Carlo sampling technique.
By using these techniques it will increases high detection rate
and providing the less false positive rate.
V. SECURE ATTACK DETECTION AND
LOCALIZATION IN WIRELESS SENSOR
NETWORKS
In this paper [4] to make a secure Localization module
which makes the particular clusters as a reference points and
by knowing the positions and by having the centroid to
provide the reference points. It must be needed to eliminate
the intrusion, the cryptographic authentication used were it
share the key which is unique, it provides malicious node
detector which is installed in beacon nodes and the regular
node, each sensor node must know about the position of the
beacon before the actual deployment. The malicious beacons
have to be filtered first due to the inconsistency among the
multiple beacon signals. It provides the Petri net provides an
Enhanced Secure Localization Scheme which protects not
only to the distance reduction attacks and also by the distance
enlargement attacks and also ensured to find the beacon nodes
which have been compromised this cryptographic
authentication and encryption does not completely prevent the
localization of the Wireless Sensor Networks (WSN). Hence
the costs of the sensors are less and the algorithms that we are
developing should be less weight. By using the triangle based
method we can detect an attack, some protocols adjust to
failures and by the multiple inclusion of the sensors which
results in the possibility of misguiding of the signals, even
though by implementing several technologies are used there
will be the compromise of the beacon nodes which shows the
different locations. Mostly sensors are static which provides
the trialteration localization method to detect the sensor nodes,
where the position of the regular node on the plane can be find
out the distances of the three different beacons, by the
availability of the resources to improve the precision of the
localization is called the multialteration. Certain
authentication schemes are needed to deploy like tesla, but
this authentication will not guarantee about the security. It is
essential to separate the beacon and compromised beacons and
since the sensor nodes are prone to errors and make them
identifies the suspicious location references. There is no need
for double check for the energy conservation the combining of
the nodes based on the secure communication. Attacks against
discovery of the location can be made to assist other types of
attacks which give the network topology; example of this is
the sinkhole attack.
VI. LOCALIZATION OF MULTIPLE MOBILE
AGENTS ON DISTRIBUTED SENSOR
NETWORK
In this paper [5] locating the mobile agents is needed
to make an interaction where the sensors are deployed on the
machines and it cannot be used for the multiple mobile agents.
Solutions are provided to locating the mobile nodes; At first
the multiple agent localization for the architectural design, by
using the RFID system combines with the target system,
implementing the localization algorithm will provide the exact
prediction of the multiple agents. Mobile agents contain the
sensor nodes by the help of these sensors which are installed
is to find out the location and the classic localization
algorithm is used by the sensors installed in the mobile agents.
It works perfectly to find out the position and the main
purpose is to improve the accuracy determination of the
location of the nodes. In the hybrid network it uses the
International Journal of Computer Science Trends and Technology (IJCST) Volume 2 Issue 1, Jan-Feb 2014
ISSN: 2347-8578 www.ijcstjournal.org Page 51
localization algorithm to find out positions inside the
particular region through which the moving objects are also
getting concerned. The node which has the control over
particular space knows the position of the mobile agent; all the
mobile agents know their positions in the network. RFID tags
stores the ID and the positions of the data through which the
infected data and the wrong readings will not occur. These
tags are stored in a particular value and the distance between
them is calculated, the position of the node is continuously
updated and the position is calculated based on the data
gathered. But the distribution of the tags affects the
localization. Tags are usable through which the correctness is
obtained and the algorithms are compared through which the
performance is measured.
VII. CONCLUSION
Machine learning approach is the different technique which
is used to find out the location of the node by giving the
Signal Strength as input and by having the already gathered
data, the position of the jamming node in the network is
difficult to find out and that can be find out by using the
Double Circle Localization. Non-Cryptographic methods are
difficult to control where the traditional methods are not
efficient and by using the Least Squares (LQ) which was used
by the multi-alteration method to point out the attacker
location, the accuracy of the algorithm results in the accuracy
of locating the attackers, Enhanced Signal Localization
scheme will be against the distance enlargement attacks and
the distance reduction attacks and results in the accuracy of
locating the attackers.

REFERENCES
[1] A.Tran, XuanLong Nguyen, and Thinh Nguyen,Localization
Algorithms and Strategies for Wireless Sensor Networks (Eds:
Guoqiang Mao and Baris Fidan, IGI Global)
[2] Zhenhua Liu, Hongbo Liu, Wenyuan Xu, and Yingying Chen
Exploiting Jamming-Caused Neighbor Changes for Jammer
Localization,IEEE Transactionson Parallel and distributed Systems.
Vol: 23, March 2012.
[3] Yingying Chen, Wade Trappe and Martin, R.P. Attack Detection in
Wireless Localization, IEEE Infocom.2007.
[4] Wen Tao Zhu Yang Xiang Jianying Zhou Secure localization with
attack detection in wireless sensor networks, Int. J. Inf. Secur. 2011.
[5] Byoung-Suk Choi, Joon-Woo Lee, Ju-Jang Lee, Kyoung-Taik Park
Distributed Sensor Network Based on RFID System, Wireless
Sensor Network, 2011.